Question

这应该很简单。我有其他PIVOT SQL查询正常工作。我想计算登录数：按小时计算，按月计算。我在想两个PIVOT或UNPIVOT，然后是PIVOT？是的，我在这里挖过，其他网站，Google等等。我很困惑。

SELECT
loginid
,DATEPART(MONTH,logtime) Month
, DATEPART(HOUR, logtime) Hour
FROM somelog (nolock)
) temp
PIVOT (
COUNT(loginid)
        FOR Month in (JAN,FEB,MAR,APR,MAY,JUN,JUL,AUG,SEP,OCT,NOV,DEC)
) AS Pvt

我想要的结果是什么..

HOUR，JAN，FEB，MAR

00

01

02

...

23

我不需要8760（365 x 24）个

我尝试过GROUP BY HOUR 我试过了GROUP BY temp.hour

我也尝试了这个..它似乎确实有效，但我得到了00小时例如365次..再次，GROUP BY问题..

SELECT
TimeOfDay
, [1] JAN
, [2] FEB
, [3] MAR
, [4] APR
, [5] MAY
, [6] JUN
, [7] JUL
, [8] AUG
, [9] SEP
, [10] OCT
, [11] NOV
, [12] DEC
FROM (SELECT logintime
, loginid
, datepart(month, logintime) [month]
, DatePart(hour, logintime) TimeOfDay
FROM sometable (nolock)) x
PIVOT (
COUNT(loginid)
for [month] in ([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12])
) AS pvt
ORDER BY 1, 2

谢谢，

肯特

Answer 1

一种方法

WITH hour_list AS (
  SELECT 0 hour
  UNION ALL
  SELECT hour + 1 FROM hour_list WHERE hour < 23
)
SELECT h.hour,
       COALESCE(jan, 0) jan,
       COALESCE(feb, 0) feb,
       COALESCE(mar, 0) mar,
       COALESCE(apr, 0) apr,
       COALESCE(may, 0) may,
       COALESCE(jun, 0) jun,
       COALESCE(jul, 0) jul,
       COALESCE(aug, 0) aug,
       COALESCE(sep, 0) sep,
       COALESCE(oct, 0) oct,
       COALESCE(nov, 0) nov,
       COALESCE(dec, 0) dec
  FROM hour_list h LEFT JOIN
(
  SELECT DATEPART(HOUR, logtime) hour, 
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  1 THEN 1 END) jan,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  2 THEN 1 END) feb,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  3 THEN 1 END) mar,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  4 THEN 1 END) apr,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  5 THEN 1 END) may,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  6 THEN 1 END) jun,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  7 THEN 1 END) jul,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  8 THEN 1 END) aug,
         SUM(CASE WHEN DATEPART(MONTH, logtime) =  9 THEN 1 END) sep,
         SUM(CASE WHEN DATEPART(MONTH, logtime) = 10 THEN 1 END) oct,
         SUM(CASE WHEN DATEPART(MONTH, logtime) = 11 THEN 1 END) nov,
         SUM(CASE WHEN DATEPART(MONTH, logtime) = 12 THEN 1 END) dec
    FROM somelog (NOLOCK)
   GROUP BY DATEPART(HOUR, logtime)
) l 
    ON h.hour = l.hour

或PIVOT

WITH hour_list AS (
  SELECT 0 hour
  UNION ALL
  SELECT hour + 1 FROM hour_list WHERE hour < 23
)
SELECT h.hour,
       COALESCE([1],  0) jan,
       COALESCE([2],  0) feb,
       COALESCE([3],  0) mar,
       COALESCE([4],  0) apr,
       COALESCE([5],  0) may,
       COALESCE([6],  0) jun,
       COALESCE([7],  0) jul,
       COALESCE([8],  0) aug,
       COALESCE([9],  0) sep,
       COALESCE([10], 0) oct,
       COALESCE([11], 0) nov,
       COALESCE([12], 0) dec
  FROM hour_list h LEFT JOIN 
(
  SELECT DATEPART(MONTH, logtime) month,
         DATEPART(HOUR, logtime) hour,
         COUNT(*) log_count
    FROM somelog (NOLOCK)
   GROUP BY DATEPART(MONTH, logtime), DATEPART(HOUR, logtime)
) s
PIVOT
(
  SUM(log_count) FOR month IN([1],[2],[3],[4],[5],[6],[7],[8],[9],[10],[11],[12])
) p
    ON h.hour = p.hour

两个查询的示例输出：

| HOUR | JAN | FEB | MAR | APR | MAY | JUN | JUL | AUG | SEP | OCT | NOV | DEC |
|------|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
|    0 |   2 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
|    1 |   1 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
|    2 |   0 |   1 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
|    3 |   0 |   0 |   1 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
|    4 |   0 |   0 |   0 |   1 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |   0 |
...

这是 SQLFiddle 演示（使用CASE）
这是 SQLFiddle 演示（使用PIVOT）

多个PIVOTS？需要按每月小时计算

1 个答案: